ebook img

Studies of instruments for measuring radiant energy in absolute value: an absolute thermopile PDF

50 Pages·2011·3.77 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Studies of instruments for measuring radiant energy in absolute value: an absolute thermopile

. STUDIES OF INSTRUMENTS FOR MEASURING RADIANT ENERGY IN ABSOLUTE VALUE: AN ABSOLUTE THERMOPILE W. W. Coblentz and W. B. Emerson CONTENTS Page. I. Introduction 504 II. Apparatus and methods 505 1. Theradiometer 505 2 Constructionofthereceivers 506 3. Thermopileandgalvanometer 511 4. Water-cooledshutteranddiaphragm 513 5. Theradiator 515 6. Theassembledapparatus 516 7. Methodofmakingobservations 518 8. Methodofreductionofdata 520 9. Correctionsfordiffusereflectionfrom thereceiver 522 10. Accuracy attainable 525 III. Experimental Data 530 ReceiverNo. 1 542 ReceiverNo. 2 542 ReceiverNo. 3 542 ReceiverNo. 4 542 ReceiverNo. 5 543 ReceiverNo. 6 543 ReceiverNo. 7 544 ReceiverNo. 8 545 ReceiverNo. 9 545 ReceiverNo. 10 546 ReceiverNo. 11 546 ReceiverNo. 12 • 547 ReceiverNo. 13 548 ,. IV. Summary 548 503 504 Bulletin 0} the Bureau 0} Standards [vu.12 INTRODUCTION I. One of the chief needs in the measurement of radiant energy is a reliable radiometer which evaluates the observations in absolute measure. Such an instrument should be a primary one, capable of measuring radiant energy directly. Because of the lack of such a primary instrument, a calibration of radiometers against a standard of radiation is advocated; and such standards of radia- tion have been prepared * by this Bureau. There are several apparently trustworthy methods for making measurements ofradiation in absolutevalue. One oftheproblems undertaken in this Bureau is the study of various instruments used in making these absolute measurements and the purpose of ; the present paper is to report on the results obtained with one of these instruments. One of the by-products of this investigation will be the evalua- tion of the coefficient or the so-called Stefan-Bolzmann constant of total radiation of a uniformly heated cavity or so-called black body. As will be shown in a subsequent paper, the different methods employed by various experimenters in evaluating this constant give somewhat discordant results. It is therefore desir- able to emphasize at the very beginning of this paper that, until itcanbeshownwhich ofthesevarious methods is themostreliable, it is undesirable to depend upon a single numerical value. In beginning the subject of radiometry in absolute measure, it seemed desirable to study instruments which might prove of use in general radiometric work (e. g., useful as radiation pyrom- eters), leaving for later investigation the more cumbersome and complicated devices for determining with the highest precision the coefficient of total radiation of a uniformly heated cavity or so- called black body.2 The first communication 3 on this subject gave the results of a study of the " Radiobalance." 4 This is an ingenious device in which the radiometric receiver consists of a thermojunction which can be heated electrically. In this manner the heat generated in the receiver by absorbing radiant energy is 1This Bulletin, 11, p. 87; 1914- sThisBulletin,9,p.51;1912. 2ThisBulletin, 12,p.553; 1916. *Callendar, Proc. Phys. Soc, London, 23, pt. 1; Dec, 15, 1910. e2SH£J Values of the Constant of Total Radiation 505 neutralized by the well-known Peltier (cooling) effect, which occurs when an electric current is passed through a thermojunc- tion. The result ofthestudy ofthis instrument showed thatwhile it has many good points it did not appear to be sufficiently reliable for a primary instrument. The investigation of the instrument used in the present work was undertaken in order to determine its reliability as a precision radiometer, i. e., to determine whether it is a primary instrument capable of measuring the radiation constants directly or whether it is a secondary instrument which must be calibrated by exposing it to a standard of radiation. The first question raised in regard to the use of a receiver of this type is the uniformity of temperature distribution when heated electrically from within" and when heated by absorbing radiant energy incident upon its front surface. From the concordant results obtained with metallic strips, differing by 10 times in thickness, and having different kinds of absorbing surfaces, it is evident that whatever errors (if any) are introduced, are much smaller than the variations observed with the various receivers constructed of the same kind of material. APPARATUS AND METHODS II. Below are given the essential features of the instruments used in this research. A more complete description of the thermopile, galvanometer, and other accessories is given in a previous paper.5 THE RADIOMETER 1. The radiometric receiver,6 Fig. 1, which was the subject of the present investigation is a very simple device, consisting of a very thin strip of metal, blackened to absorb radiation. Behind this metal strip, at a distance of 2 to 3 mm, is placed a thermopile of bismuth-silver having a continuous receiving surface of tin. The thermopile is connected with an ironclad Thomson or a d'Arsonval galvanometer which serves merely as a null instrument to indi- cate the rise in temperature of the metal strip. 5This Bulletin, 11, p. 132; 1914. 6Forfurtherdetails,seethisBulletin, 11,p. 157;1914. — 5°6 Bulletin of the Bureau of Standards [Vol.12 CONSTRUCTION OF THE RECEIVERS 2. — The strip of metal serves three purposes (i) as a receiver for absorbing radiant energy, (2) as a source of radiation which can be produced by heating the strip electrically, and (3) as a standard of radiation to test the galvanometer sensitivity, by heating the strip electrically by a standard current. The use of a thermopile, separated from the receiver, makes it possible to use various receivers. The radiometric apparatus was 4:8CM Fig. I. Theradiometer therefore designed so that the receivers could be mounted over the thermopile as shown in Fig. 2. Each receiver, as shown in Fig. 2, B is a complete unit consisting of an insulating base of slate (cut mm from an ordinary writing slate), 3 in thickness, to which are E mm attached the copper electrodes, E, 0.5 in thickness, the receiver R, the potential terminals P P, the knife-edge slits 5 S, at the front, and the strips of copper, C C, at the rear, which are used for the purpose of preventing radiations, not intercepted by — Coblentz"I Values of the Constant of Total Radiation 507 Emerson] thereceiver, fromfallinguponthe thermopile. Theplatinum used for receivers was the "platinum in silver" material used for bolometers. This material was cut into strips, by means of a specially constructed cutting device, and soldered to the elec- trodes, which were then covered with Chatterton compound in order to protect them from the nitric acid which was used in mm t=o.5 Rear View op Receiver IKt-l.<7mm End Section through Aft Fig. 2. Showingtheconstructionofthereceivers removing the silver from the platinum. The platinum strip was then cleaned electrolytically by dipping it in hydrochloric acid. After attaching the potential wires, the strip was coated electro- lytically with platinum black.7 Incidentally, it may be added that the platinum cholride seems to deteriorate rapidly with usage, giving grayish deposits, so that it is necessary to use freshly prepared solutions. 7ThisBulletin,9,p.305;1913. Kurlbaum,Ann.derPhys.(3),67,p.846;1899. 508 Bulletin of the Bureau of Standards [va.12 The potential leads P P consisted of No. 36 silk-covered copper wire attached to the holder by means of Chatterton compound. To the end of this copper wire was soldered a fine platinum wire mm m 0.025 thickness, and (for several receivers) to the end of this platinum wire was soldered a short piece of (Wollaston) platinum mm wire about0.01 in thickness. To the free end ofthe latterwas mm attached a bead of solder from 0.01 to 0.03 in diameter, and this was melted to the platinum strip by means of the small nichrome 8 heater now used in performing such delicate opera- tions. If thebead of solder is dipped in a solution of zinc chloride, there is no difficulty in attaching it to the platinum strip, and the mm juncture will be only 0.05 to 0.1 in diameter. This juncture is covered with shellac to permit repairs in case of breakage, and to protect the solder from the-platinum chloride solution when depositing the platinum black. These potential terminals were mm situated at a distance of 3 to 4 from the copper electrodes. In order to test the effect of the potential contacts, an additional wire, c, was attached to two of the receivers. The platinum potential wires are shellacked for insulation. The opening in front of the platinum receiver is defined by means of knife-edged strips of metal, along the sides and across the ends. The latter were placed directly over the potential terminals. When it was found impossible to detect a difference in the radiation constant, as measured with and without these slits (which were made of thin copper) across the ends, they were discarded and the entire length of the receiver was exposed to radiation. This eliminated the conduction of heat from the region between the potential terminals, which were then used to define the length of the receiver. Knife-edge slits 5 5 were used over the sides of the receiver, in order to define its effective width which is difficult to determine after the receiver has been covered with platinum black; and, especially so, after smoking it with mm lampblack when a layer of soot 0.05 in width may adhere loosely along the edge. These knife-edged slits were of two mm kinds: (1) Brass ones 0.5 in thickness, and (2) slits of alumi- mm num 1.95 in thickness. Both types were bright over a width 8ThisBulletin,9,p.7; 1912. £SSJ Values of the Constant of Total Radiation 509 of about 1 cm along the knife-edge, on the side exposed to the radiator. Measurements were made without the slits and the value of the radiation constant was usually slightly higher than with the slits. In view of the possibility of diffusely reflected radiations from the radiator reaching the thermopile, the slits were used. It would have been desirable to use a hemispherical mirror in front of the receiver, as in previous work,9 and thus eliminate the correction for the loss of radiation by diffuse reflec- tion, but the uncertainty then arising as to the exact area exposed and as to numerous adjustments, etc., appeared to introduce greater errors than would arise from a lack of proper correction for reflection, which correction was determined directly. The re- ceiver of the thermopile had an area of 1.8 by 19 mm, the length being somewhat less than the distance between the potential terminals on the receiver R. The thermopile was covered perma- mm nently with apiece ofcardboard, about0.4 in thickness, which mm mm had an opening, about 2.5 in width and 20 in length, to admit radiations from the receiver. This cardboard shield was used to exclude the radiations coming from that part of the receiver which extends from the potential terminals to the copper electrodes. In addition to this shield, each receivermountingwas covered on the rear side with heavy (0.8 mm) strips of copper (or sheet iron folded as in receiver No. 6), C C, Fig. 2, to protect the thermopile from possible stray radiations which might not be intercepted by the receiver. These strips of metal were painted on both sides with lampblack, then blackened by holding them in the flame of a sperm candle. This method of construction places mm the thermopile surface at a distance of about 3.5 from the receiver, which greatly reduces the sensitivity. However, it eliminates the question of convection currents produced by the warming of the receiver. In the first part of the work the shields, C C, were not used, and consequently the thermopile was much closer to the receiver, which increased the sensitivity. The same numerical value of the constant of radiation was obtained as with the greater separation of the receiver and the thermopile. The method of blackening the receivers with soot is of some importance. The thick metal plates used as shields were painted 9ThisBulletin,10,p.2; 1913. 510 Bulletin of the Bureau of Standards [Voi.12 with lampblack and smoked by drawing them through the tip of the flame of a sperm candle. This produces a deposit of soot which reflects but little more than 0.5 per cent.10 However, as indicated in the previous paper relating to this subject, the cold deposits of soot from an acetylene flame or sperm candle reflect a little over 1 per cent. In order to obtain the blackest deposit from a sperm candle, the wick must be free from old, charred material which produces a "hard" bluish smoke, having a high reflecting power. On the other hand, the wick must not be trimmed too short. The best results are obtained by using a freshly trimmed candle in which the flame is burning at its nor- mal height. The smoke is produced with a sheet-iron cone about 8 cm long and 4 cm at its base, flattened at the top, leaving mm an opening 1.5 by 10 for the escape of the soot. When the candle is burning properly this funnel is held over the flame by means of a crucible tongs, and the object which is to receive the soot is passed back and forth, at a distance of 2 to 5 cm, over the top of the metal cone, which is not permitted to become hot. As to the thickness of the deposit of lampblack, that seems to be a matter of guesswork. The paint is a mixture of lamp- black in alcohol and turpentine which is thoroughly shaken and allowed tp stand a few minutes so that the coarse agglomer- ations may settle to the bottom. The coating of paint applied to a receiver is of such thickness that in bright sunlight the metal underneath is barely perceptible through the thin spots in the paint. This surface of lampblack paint is then exposed in the smoke of a sperm candle until it is thoroughly covered with mm soot. Some layers of soot were, no doubt, as much as 0.1 in thickness, although the deposit is usually just sufficient to cover the paint. The receivers made of manganin or therlo and painted with lampblack, Fig. 3, do not have an appreciable (to 1 part in 2650) temperature coefficient of resistance. On the other hand, the temperature coefficient of resistance of platinum, which is positive when unblackened, becomes negative after it is given a coating 10ThisBulletin,9,p.283; 1913. — Coblentz1 Values of the Constant of Total Radiation 5ii Emerson] of platinum black. This is illustrated in Fig. 3, which shows the change in resistance of the strip when heated to different temperatures, as 'indicated by the current which was passed through the strip in order to raise it to the same temperature as was produced by the absorption of radiant energy. .0t8 -OSOTherlO .Off .C^3pT Fig. 3. Showing the change-inresistance with temperature {heatingcurrent)of receivers oftherlo, No. j, andplatinum THERMOPILE AND GALVANOMETER 3. These two instruments have been described elsewhere,11 and it will be sufficient to add that the thermopile with its auxiliary receiver required, at a minimum, from 12 to 15 seconds (depend- ing upon the receiver used) to produce a maximum galvanometer deflection. This long period requires uniform conditions in order to obtain accurate measurement. No observations were made, therefore, on very windy days when the thermopile was affected 11ThisBulletin,11,p.132; 1914. 512 Bulletin of the Bureau of Standards [Voi.iz by air currents. The sensitivity was such that it would have been possible to observe the rise in temperature of the thermopile by means of a d'Arsonval galvanometer. Two such instruments were available, but they happened to respond to vertical tremors in the laboratory so that readings could not be made with an mm accuracy higher than 0.2 in a maximum deflection of 60 to 70 mm. A deflection of this size (75 per cent larger than normal) was obtained by reducing the critical damping resistance of 100 ohms to 50 ohms. This increase in sensitivity was, of course, at the expense of the period. The d'Arsonval galvanometer was therefore used simply as a check on the observations made with an ironclad Thomson galvanometer of 5.3 ohms, which was operated on a single swing of 1.5 to 2 seconds. But even on this short period the instrument was too sensitive, so that an external resistance of 60 .to 120 ohms was constantly used in the circuit. The galvanometer deflections were then from 150 to 200 mm, and under these conditions it was the usual experience to repeat the readings to 0.3 to 0.5 mm. The galvanometer being completely shielded magnetically, the unsteadiness observed was found to be caused by air currents in the thermopile. On rare occasions the galvanometer was momentarily deflected, which was attributed to radiotelegraphic disturbances. Great steadiness in the galvanometer is necessary because of the long wait (about 30 seconds at a minimum) to obtain the complete scale reading. No difficulty was experienced with the galvanometer, and in the present work plenty of time could be allowed for the receiver and the thermopile to attain a steady temperature. Moreover, after the deflection had attained a maximum it would stay at that point whether the stimulus was applied for 20 or 30 seconds or for a much longer period. This is an important point in radiometric work of this type. The gal- vanometer was not completely damped on a swing of less than 2.5 seconds, and, because of the quickness of the response of the thermopile, especially when the receiver was of thin bolometer platinum, the maximum steady deflection was attained by exe- cuting one large throw and one or two small vibrations before the needle came to rest. This large first throw of the galvanometer

Description:
One of the chief needs in the measurement of radiant energy is a reliable . is covered with shellac to permit repairs in case of breakage, and to protectthe radiations from the radiator reaching the thermopile, the slits were used.
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.